| |||||||||||||
why did RMS decide not to honor his commitment to "do no harm?why did RMS decide not to honor his commitment to "do no harm?Posted Oct 2, 2006 4:33 UTC (Mon) by shieldsd (subscriber, #20198)Parent article: Busy busy busybox
This BusyBox business is just one more sign of the abyss we are facing. We all know Linus was heavily tied to GPL2. And RMS promised to "do no harm" per section 1.3 of the process document; so the proper approach would have been for RMS and Linus to work together in a series of license "patches" to avoid the possibility of forking; that is, refine the license in the same way as the code is developed. This would have allowed full community participation at every step.
Yet RMS decided to pull ahead on his own and added the DRM provisions. This made forking almost inevitable, which means we are going to see more projects forking. Each such fork will divert developer resources needlessly, thus slowing the rate at which the kernel and associated software packages can grow.
It's a shame, and it could have been avoided.
I've some more thoughts on this at my blog: http://daveshields.wordpress.com
(Log in to post comments)
Who is doing harm here? Posted Oct 2, 2006 5:05 UTC (Mon) by atai (subscriber, #10977) [Link] The GPL v3 process is open (and open to all). Linus has his avenue of commenting, like anyone else. Instead all these "mine" not accommodated, not at "my" convenience, not letting "me" seeing the first draft first, etc. (where me, mine, my is Torvalds) Who is really doing harm here?
There is an year for making comments. Yet Torvalds just sit there without adding his comments. And he said others know his email address. Is it hard for him to add some text to a web form? Instead he flames with PJ and makes silly statements to the press. Making a whole paragraph belittling the FSF over a simple clarification. He shows no willingness of direct communication and it is either Linus' way or the highway.
The FSF has shown more maturity in the process than Linus show. Linus is, honestly, behaving childishly.
Who is doing harm here? Posted Oct 2, 2006 6:05 UTC (Mon) by zonker (subscriber, #7867) [Link] I have to disagree here. I belive Linus' point, or at least one of them, is "who wanted the GPLv3 inthe first place?" Participating in the GPLv3 process is essentially saying "ok, I've decided to let the FSF set the terms of the debate and the new license, and will simply be content to quibble over the particulars of the framework that has been handed down by RMS." I am not at all surprised that Linus, and other interested parties, have opted to sit out the GPLv3 process when it was never established that these people wanted the license revised in the first place.
Who is doing harm here? Posted Oct 2, 2006 7:09 UTC (Mon) by khim (subscriber, #9252) [Link] They were involved in discussion from the moment when Linus said I think it's insane to require people to make their private signing keys available. But it was just some internal discussion so it's Ok. While kernel guys sat out and just ignored GPLv3 - I was content. It's their right. But when they are starting to play politics - they are clearly involved and then the question arises: why have you ignored everything when you were asked but when process is nearing finish you are raising tempest ? Sorry guys - if you want to discuss do it in the proper way. If you want patch to be included in linux kernel - you sent it on LKML, not publish it on corporate website and then complain that "this @%#!@^&*#!#@ Linus just refuses to grab our wonderfull creation - and he knows where our web site it".
Who is doing harm here? Posted Oct 2, 2006 7:46 UTC (Mon) by atai (subscriber, #10977) [Link] OK, if Linus never wants v3, then fine. Just say Linux is GPL v2 forever and be done with it.
There are people who want GPL v3. There are people who want the DRM threat addressed. There are people who want the ASP loophole closed. There are people who want better license compatibility. The kernel developers may think GPL v2 is fine for them, but they do not consider the needs of the people caring about the ASP scenario, for example.
Linus is now attacking the GPL v3 beyond what he wants to do with the kernel (staying with v2). He can just ignore the GPL v3, but the recent open letter from the kernel developers and Linus's statements make you wonder if he is trying to derail the use of GPL v3 outside his project in general. And he seems happy playing this role (due to his personal hatred toward RMS?)
Who is doing harm here? Posted Oct 2, 2006 8:34 UTC (Mon) by dlang (✭ supporter ✭, #313) [Link] you want greater license compatability by creating a license that's incompatable with all the existing GPLv2 code???!!?!?!?
you are asuuming that all the people who licensed code as GPLv2 or later are going to be willing to change that to GPLv3 or later.
if they want to stick to the GPLv2 or later then they have to live with the license compatability limitations of GPLv2
as documented elsewhere in this thread, this opens the possibility of a three way fork of a project
fork 1
with these three forks the only code 'shareing' that could be done is that fork 2 could get code from fork3. any other shareing would involve the author needing to authorize a license change.
yes fork2 can also take code from projects with more licenses, but those other projects can't take code back from the GPLv3 project without changeing the license of their own code (and if they wanted to do that they could anyway, without needing this 'compatable GPLv3 license' simply by relicensing their code
so exactly how is this 'improved license compatability' supposed to improve code shareing between projects? everything looks one-way to me.
Who is doing harm here? Posted Oct 2, 2006 20:44 UTC (Mon) by stevenj (guest, #421) [Link] you want greater license compatability by creating a license that's incompatable with all the existing GPLv2 code???!!?!?!? The poster was obviously referring to section 7 of the GPLv3, which makes the new GPL compatible with more licenses than the old GPL, most notably the new Apache license. Anyone who made their code "v2 only" knew that they would be incompatible with future GPL versions, just as GPLv2 was incompatible with GPLv1. I fail to see how you can criticize the FSF for something that was impossible to avoid without freezing the GPL forever at v2. And the history of free software is rife with FUD about forking that hasn't lived up to reality (even with licenses like LGPL that theoretically allow legally incompatible forks). Malign forking doesn't happen (in the FLOSS world at least) unless the project already has serious problems unrelated to licensing.
Who is doing harm here? Posted Oct 2, 2006 21:09 UTC (Mon) by stevenj (guest, #421) [Link] with these three forks the only code 'shareing' that could be done is that fork 2 could get code from fork3. any other shareing would involve the author needing to authorize a license change. By the way, your conclusion is false on the facts. The only code that could not be legally combined in your hypothetical example is that fork 1 ("v2 only") could not share code with fork 2 ("v3 or later"). It would be perfectly legal to combine code from fork 1 ("v2 only") and fork 3 ("v2 or later") or from fork 2 ("v3 or later") and fork 3 ("v2 or later"). There might be social obstacles, but there are no legal obstacles (no legal authorization from the authors is required). Any any hypothetical project whose developers are so divided that they undergo a triple fork has more problems than just licensing.
Who is doing harm here? Posted Oct 2, 2006 11:36 UTC (Mon) by smitty_one_each (subscriber, #28989) [Link] >say Linux is GPL v2 forever and be done with it.
If "done" were possible.
Who is doing harm here? Posted Oct 2, 2006 19:57 UTC (Mon) by landley (guest, #6789) [Link] > OK, if Linus never wants v3, then fine. Just say Linux is GPL v2> forever and be done with it.
Friday, September 8, 2000 was six years ago:
That predates the entire GPLv3 debate by a lot, and it isn't Linus who's
Who is doing harm here? Posted Oct 2, 2006 7:15 UTC (Mon) by error27 (subscriber, #8346) [Link] PJ is a nice person, but she's also wrong. GPLv3 sucks.
I don't see how the FSF can keep claiming that Linus hasn't explained his objections clearly. He may not have gone through all the formal process, but they must be vegitative to claim he isn't communicating his thoughts. You seem to be implying that the FSF has no one on staff who can access the internet and read.
Personally, I don't like the optional web clauses in the GPLv3. The Gnu FDL has the same optional stuff where you never know what you're getting. It caused real problems for Debian.
There may be other things that suck with the GPLv3. I'm not very familiar with it yet. How do busy box rules affect Tivo DRM, for example? TiVO DRM is at a lower level so it shouldn't matter.
Who is doing harm here? Posted Oct 2, 2006 8:34 UTC (Mon) by khim (subscriber, #9252) [Link] They must be vegitative to claim he isn't communicating his thoughts. Sorry, but he's not. First Linus claims that GPLv3 restricts the "use". FSF explains that it's not so: you are free to use (== compile and run) GPLv3 code. Linus says that it's not "use" normal people are talking about - it's "use in other projects" he's concerned with. FSF retorts that GPLv1 and GPLv2 always restricted such "use" and this is what Linus itself praising in GPLv2 ! Linus then start to talk gibbegish and claim that semantic is not important and intent is important. Thus no, Linus never explained what he does not like about GPLv3 clearly - he finds new and new ways to change his position over time... At least this text is readable - but it's hardly acceptable: it still does not explain why the "freedom to take everything, add extenstions and sell binary blobs" (aka "BSD license") is bad while "freedom to sell binary blobs with corresponding code which is useless without the hardware to run in on" is essential. Or why they think "while we may argue forcefully for our political opinions, we may not suborn or coerce others to go along with them" - this is what linux developers are doing, after all (if you listen to SCO then you'll know that "tit for tat" enforced by GPLv2 is also "political opinion"). GPLv3 does not forbid DRM. It does not even force to stop using GPLv3 code in system with DRM (after all DRM scheme can be implemented in seperate domain - and it's actually the only way to make it work). Additional restrictions is no better and no worse then LGPL - and LGPL is important license to have. Patents... Patents are discussed with IBM (the biggest paten holder of them all!) and if IBM will be satisfied (big if, but at least IBM is working on that and not starting political campaign), then how kernel developers can claim that it's a problem ? In short: it looks like Linus and kernel developers are intentionally choose the ways of communication which can be used to "poke" FSF and then go away thus avoiding any constructive discussion. Was it because they don't have time to defend their position or have they no arguments to actually defent it ? We'll see...
Who is doing harm here? Posted Oct 2, 2006 11:51 UTC (Mon) by smitty_one_each (subscriber, #28989) [Link] >Linus and kernel developers are intentionally choose the ways of communication which can be used to "poke" FSF and then go away thus avoiding any constructive discussion.
Yes, the Discussion Rights Management is every bit as contentious as the DRM of the GPLv3.
Who is doing harm here? Posted Oct 3, 2006 22:50 UTC (Tue) by man_ls (subscriber, #15091) [Link] In short: it looks like Linus and kernel developers are intentionally choose the ways of communication which can be used to "poke" FSF and then go away thus avoiding any constructive discussion.Indeed. Just as they complained that the FSF had already decided the contents of GPLv3 before starting the public consultation, it seems that kernel devs had already decided their opposition before seeing the first draft. It is not surprising, since they were already sticking to GPLv2 since the beginning. It almost seems like they want to retroactively be right: first they said: "I want to stick to GPLv2", now they are saying: "see? We was right to stick to v2, GPLv3 sucks". Meanwhile Linus was playing the prima donna complaining that "Eben [Moglen] said he was willing to go over the draft point-for-point after it was published, but not apparently willing to let me read it in the comfort of my own home and send comments back. So, the GPLv3 process couldn't apparently accommodate me, or I couldn't accommodate them."Why did he need to go first? Linus could certainly read the draft in the comfort of his own home after it was published; we all did. Of course we are not the leader of (arguably) the most important libre software project out there, but then he was talking to the lawyer of the guy who has (and I quote Ingo Molnar freely) "unprecedented legal power over 350 million lines of code" (conservatively estimated). I would expect a bit of humbleness from him. Well, what do I know. Was it because they don't have time to defend their position or have they no arguments to actually defent it ? We'll see...Ingo Molnar is displaying a lot of patience, and also spending a lot of time defending the position of the most prominent kernel devs'. He also has a lot of arguments, but so far they are not very convincing, not for me at least. They shift too much.
Who is doing harm here? Posted Oct 5, 2006 12:58 UTC (Thu) by wookey (subscriber, #5501) [Link] Indeed. I would like to commend Ingo and stevenj for their contributions to this thread, which have clarified a number of things for me at least. The whole thing has been pretty civil as these things go. I do now at least understand a bit of why the kernel devs are so grumpy.
But I do not find myself convinced by their arguments, and agree with stevenj that the fears of forking, 'legal strongarming', and 'acting with legal but not moral authority' are overblown.
Nevertheless, Jon Corbet put it well (as ever) in the original article. This debate _is_ causing strife, and there is thus quite a lot to be said for a v3 that only has the uncontentious stuff in it. That would be a pity - the FSF are the ones who are being consistent and 'in the spirit of', and the kernel devs who are being awkward (IMHO), but it is quite finely balanced as to which course would be better in the long term.
We certainly won't be able to say that the issues have not been discussed fully at the end of it all, whatever happens - and that of course, is a good thing - tiresome as it might seem right now.
So, to re-iterate. Thank you for those making an effort to dicuss the substnative issues (which is nearly everyone here). It is appreciated.
Who is doing harm here? Posted Oct 9, 2006 17:01 UTC (Mon) by malor (subscriber, #2973) [Link] In my opinion, the difference is that the FSF is thinking about freedom in a multi-generational timeframe, and the kernel devs want success/market dominance *right now*. And they appear willing to trade off freedom to get it.
If you think about it from the perspective of tens or hundreds of years, Linux is not that important; it's essentially the early scaffolding for a great cathedral of code. But, unlike the real multi-hundred-year cathedral projects, I don't think these scaffold builders are even aware that a cathedral will ever exist.
The FSF is right not to bend on this issue; if we want truly free computers, it's incredibly important. Eventually, GPLv3 code will be such an overwhelming presence that it would be insane to try to compete with it in any but the smallest niches.
If, however, the foundation is flawed, the cathedral may collapse. The architect has identified flaws in the design, and is redrawing the blueprints. The scaffold builders are upset about this. Either the cathedral won't look quite how they thought, or they think their project IS the cathedral.
The most popular free kernel, a hundred years from now, may even be called Linux, but it will be related to today's kernel only in the sense that modern man is related to the chimpanzee. But the GPLv3 may very well still be in force. Law, especially good law, lasts a LOT longer than code.
why did RMS decide not to honor his commitment to "do no harm? Posted Oct 2, 2006 5:24 UTC (Mon) by khim (subscriber, #9252) [Link] It's a shame, and it could have been avoided if "patch" will only be included when all interested parties agree. Sorry, but this will never happen. "All interested parties" will never agree because any change will bring objections from someone. And that's not how Linux kernel development works and you know it: what Linus said goes. If we have two USB stacks but Linus does not like them both - then USB stack written by Linus is included in kernel. No consensus. And the result is that some projects (especially embedded ones) are using 2.4 kernel still - because they disagree with changes introduced in 2.6 process. Sounds familiar ? Yup. Because it is. Yet RMS decided to pull ahead on his own and added the DRM provisions. ...just like Linux wrote his own version of USB stack when he was unsatisfied by existion ones. Yup. RMS took a look on TiVo and found out that a user who needs changes in the system is no longer free to make them himself, or hire any available programmer or company to make them for him and since it was the deal from the very start he decided to fix it. Linus claims that this crusage is not important and we shuld keep "tit-for-tat" principle embedded in GPLv2. The problem is: GPLv2 was never about "tit-for-tat" (Google, IBM or even "evil empire" Microsoft got to keep changes private as long as they keep binaries private). And the FSF was never about "tit-for-tat" - they clearly explained what the goal is many years ago.
|
|||||||||||||